Basis for evaluation
Central government debt management 2015
Dnr 2016/79 22 February 2016
The year in brief 1
1 Objectives and evaluation 2
1.1 Mandate 2
1.2 Objectives and governance 2
2 Debt management strategies 5
2.1 Liquidity and infrastructure 5
2.2 Transparency, predictability and clear communication 6
2.3 High confidence in the Debt Office 8
3 Cost and risk 9
3.1 Costs 9
3.2 Risks 11
4 Borrowing requirement forecasts 15
5 Summary of borrowing 17
6 Capital market borrowing 18
6.1 Government bonds 18
6.2 Inflation-linked government bonds 19
6.3 Foreign currency bonds 21
7 Money market borrowing 23
7.1 T-bills 23
7.2 Commercial paper 24
7.3 Liquidity management 24
8 On-lending of foreign currency 26
9 Derivative instruments 27
9.1 Interest rate swaps 27
9.2 Basis swaps 28
9.3 FX Forwards 28
10 Positions 29
11 Retail market borrowing 30
Basis for evaluation 2015
The cost of the central government debt fell as a result of the general decline in interest rates.
During the year the Debt Office sold government bonds at an average yield of 0.48 per cent, a decrease from 1.35 per cent in the preceding year.
See chapters 3 and 6.
The Debt Office sold government securities at a negative interest rate.
For short maturities central government was paid for borrowing. The government bond maturing in March 2019 was sold at an average yield of –0.16 per cent. The Debt Office also borrowed at a negative rate through T-bills and commercial paper.
See chapters 6 and 7.
Two new lottery bond loans were sold at a loss since the interest rate on corresponding govern- ment bonds was negative. The Debt Office’s other savings product for private individuals, National Debt Savings, was wound up as planned in 2015.
See chapter 11.
Poorer liquidity in government securities.
Market actors consider that liquidity in the market for government securities deteriorated during the year. Regulations, the low interest rate and the Riksbank's purchases of government bonds are reasons given. See chapter 2.
The request for the Debt Office’s market commitments increased in pace with purchases of government bonds by the Riksbank. The demand for the Debt Office’s repos of government bonds rose in the second half of 2015.
See chapter 6.
The central government budget deficit was more than halved. The net borrowing requirement, i.e. the deficit in the central government budget, was SEK 33 billion compared with SEK 72 billion in the preceding year. Strong growth in Sweden led to unexpectedly high tax payments. See chapters 4 and 5.
The Debt Office amortised the foreign currency debt. As a step in the Government’s decision to gradually reduce the foreign currency exposure of the central government debt, the Debt Office bought JPY 318 billion, corresponding to about SEK 22 billion. See chapter 3.
The year in brief
In 2015 the Riksbank (the Swedish central bank) reduced its repo rate from zero to –0.35 per cent and announced that it intended to buy government bonds with a nominal value of SEK 200 billion. As a result, longer rates fell too. The borrowing cost for government bonds fell to such low levels that the issuance of lottery bonds could no longer contribute to reducing the cost of the central government debt. The purchases of government bonds by the Riksbank may also have contributed to the deterioration of liquidity in the market for Swedish government securities.
1.1 Mandate
The Swedish National Debt Office (the Debt Office) borrows money on behalf of central
government, chiefly by issuing government bonds in the Swedish and international capital market. These loans finance deficits in the central government budget. When the budget shows a surplus, the Debt Office amortises the central government debt.
In addition to continuously handling surpluses and deficits in the central government budget the Debt Office raises loans to replace loans that mature.
This means that old deficits in the central government budget are refinanced.
So the Swedish and international investors who buy Debt Office bonds are lending money to the Swedish State. These investors may, for example, be insurance companies, banks, investment funds and central banks.
A small part of the central government debt is financed by private individuals, chiefly through lottery bonds.
1.2 Objectives and governance
The overall objective for the management of the central government debt is to minimise the cost of the central government debt in the long term without the risk becoming too high. In practice this means first setting the level of risk that can be accepted in the management of the debt. This is followed by the determination of the cost of the prevailing market prices when each loan is raised.
At a general level the Government governs the management of the debt by adopting Guidelines for central government debt management each year. The basis for that decision includes the Debt
Office’s proposed guidelines. The government guidelines determine, for example, the distribution of the debt between nominal krona debt, inflation- linked krona debt and foreign currency debt. The government guidelines also regulate the maturity of the different debt types.
The guidelines limit the exposure of the central government debt to different risks. However, the guidelines say nothing about financing as such, i.e.
the types of debt and the maturities actually used by the Debt Office in its borrowing. The Debt Office uses derivative instruments so as to be able to borrow in a flexible and effective way while keeping risks within the limits set out in the guidelines.
The government guidelines provide that the Debt Office is to adopt internal guidelines. They are intended to regulate the composition of the foreign currency debt, principles for market and debt maintenance and certain other matters. The Board of the Debt Office adopts these internal guidelines in the Financial and Risk policy.
The Debt Office’s main contribution to minimising the cost of the central government debt is to work to make the market for Swedish government securities attractive to as many investors as possible. If many investors want to lend money to the Swedish State, the cost of the central
government debt will be lower. This also promotes liquidity in the market, which then stimulates demand for government securities. Another aspect is that the financing risk is lower if many lenders are prepared to lend money to the Swedish State.
At the margin the Debt Office can also influence the cost of the central government debt by making use of occasions when the pricing in the market is favourable.
1 Objectives and evaluation
The overall objective is to minimise the cost of the central government debt in the long term without the risk becoming too high. The Debt Office has also identified a number of intermediate objectives that are easier to influence and measure. This chapter describes how the management of the central government debt is governed and the possibilities of evaluating the management of the debt.
Background to the current guidelines
Shares of different types of debt
The purpose of borrowing in several different types of debt is to be able to attract a broad range of investors and spread risks. This lowers the aggregate risks in the debt management.
Nominal krona debt
The most important type of debt is nominal government bonds. Here the Debt Office has an important task to promote the functioning of the market and market infrastructure, partly by meeting the need for bonds with different maturities.
The Debt Office also issues T-bills in order to borrow at short maturities. Along with short-term borrowing in foreign currencies (see below), the T-bills are mainly used to cope with seasonal variations in the borrowing requirement.
Inflation-linked krona debt
By issuing inflation-linked krona bonds the Debt Office can attract investors who want to avoid the risk of inflation eroding the value of their bonds.
Foreign currency debt
In the international capital market the Debt Office is able to borrow large amounts in a short space of time. So there are reasons for borrowing regularly in foreign currency even when the borrowing requirement is small since this maintains the preparedness to borrow large amounts if needed.
In a recent reassessment of the cost and risk of foreign currency debt, the Debt Office has not been able to demonstrate that currency exposure reduces costs in the long term. However, it adds to the cost variation. The Government has therefore decided to gradually reduce the currency exposure. This means that, to a greater extent that in the past, foreign currency
borrowing will be converted into Swedish kronor by use of derivatives.
Maturity of the central government debt Historically, short (floating) rates have been lower than long (fixed) rates, so it has usually
been cheaper to borrow in short than in long maturities. On the other hand, the cost of the central government debt varies more with short- term borrowing since the interest rate is altered more frequently.
The maturity is decided by weighing the benefit of minimising cost (borrowing at floating interest rates) against the benefit of having low cost variation (borrowing at fixed interest rates). In the past ten years the maturity of the central
government debt has been between three and four years measured as duration.
For some time now the costs of short-term and long-term borrowing appear to have moved close to one another. So the advantage of short-term borrowing has decreased. The Government has therefore decided to extend the maturity of the central government debt.
Practical aspects
The choice of maturity is also influenced by other aspects. For example, it is in the interest of the Debt Office to ensure that all bond loans in the Swedish krona market, whatever their maturity, are large enough for bond trading to function well.
Well-functioning derivatives markets facilitate the steering of the maturity of the central government debt. In the nominal krona debt the Debt Office is able to allow the maturity to deviate from the maturity of the bond loans by using interest rate swaps. However, the Debt Office has made the judgment that it does not want to be too big a participant in the Swedish swap market.
For the inflation-linked krona debt there is, in practice, no possibility of using derivative instruments to steer maturity. Therefore the maturity of the inflation-linked krona debt is steered solely by the maturity of the bonds.
In foreign currencies there are well-functioning markets for interest rate swaps and here the Debt Office is a relatively small participant. The maturity of the foreign currency debt can there- fore be steered without influencing the market.
Government guidelines for 2015
According to the government guidelines for 2015 the central government debt is to have the following composition:
the share of inflation-linked krona debt in the central government debt is to be 20 per cent in the long term.
the foreign currency exposure of the central government debt is to decrease by no more than SEK 30 billion per year.
the remaining part is to consist of nominal krona debt.
According to the government guidelines for 2015 the maturity of the various types of debt, measured as duration, is to be
Foreign currency debt: 0.125 years
Inflation-linked krona debt: 6 – 9 years
Nominal krona debt:
Instruments with a maturity of up to twelve years1
(until and incl. 11 March) 2.3 – 2.8 years (as of 12 March) 2.6 – 3.1 year
For instruments with maturities of more than 12 years, the long-term benchmark for the outstanding volume is to be SEK 70 billion.
Under the government guidelines the Debt Office may take positons in the market to reduce the cost of or risk in the central government debt. Positions in foreign currency may not exceed SEK 300 million, measured as daily Value-at-Risk with a 95 per cent confidence level. Positions in the krona exchange rate are limited to a maximum of SEK 7.5 billion.
Evaluation and intermediate objectives Under the guidelines the evaluation of the
management of the debt is to be carried out in the light of the knowledge available at the time of the decision. However, evaluating the overall objective of cost minimisation is a complicated matter. The cost depends on the prevailing market interest rates, and the government securities offered by the Debt Office influence these interest rates. To know whether the Debt Office did minimise the loan cost
1 As a result of the sharp decrease in interest rates around the end of 2014 and beginning of 2015 the Board of the Debt Office proposed an increase of 0.3 years in the maturity interval on 18 February 2015. The Government made a decision in accordance with the proposal of the Debt Office on 12 March 2015.
it would, in principle, be necessary to know what costs alternative borrowing strategies would have resulted in.
So it is difficult to set quantitative objectives for central government’s borrowing cost. Nor is it obvious that comparisons with other borrowers give good answers. The Debt Office borrows more cheaply than any other borrower in the Swedish capital market, and compared with many other borrowers it also has low costs for borrowing in foreign currencies (see section 6.3).
As already mentioned the Debt Office works to make the government securities market as attractive as possible for investors. To concretise this work the Debt Office has broken down the overall objective into several intermediate objectives. They support the overall objective but are easier to influence and measure.
The most important intermediate objectives are that
the market in government securities is liquid and has well-functioning infrastructure
Swedish government securities attract a broad range of investors
the Debt Office has transparent and open communication
the Debt Office is clear and predictable
the Debt Office has good counterparty and investor contacts
the Debt Office achieves lower cost or risk in the central government debt by taking derivative positions in foreign currencies.
The Debt Office uses several management
strategies to achieve its intermediate objectives. An annual questionnaire survey in which primary dealers and investors are asked to rate the outcome and importance of the Debt Office's strategies and activities is one means used to evaluate how well the Debt Office is fulfilling these objectives.
The Debt Office’s management strategies are described in more detail in the next chapter. The results of the latest survey are also reported there.
The result of position taking is reported in chapter 10.
The Debt Office’s management strategies can be divided up into two categories. The first consists of the strategies that support liquidity and infra- structure in the government securities market. The second covers those that promote transparency, predictability and clear communication.
Since 2004 TNS Sifo Prospera (called Prospera below) has conducted a questionnaire survey each year among the Debt Office’s primary dealers and investors.2 The respondents were asked to rank the importance of the various strategies (see section 2.3) and to give grades on a five-grade scale as to how well they consider that the Debt Office is implementing the strategies.
Prospera also makes an overall assessment in which all the grades are weighted on the basis of the perceived importance of the strategies. Figure 1 shows the development of the overall assessment over time.
Figure 1 Overall assessment
The grade given by Swedish investors went down compared with 2014, mainly because the investors wanted to have more frequent contacts with the Debt Office. Foreign investors and primary dealers gave a slightly higher grade instead. Apart from a slight fall in 2010 the grades have been around four over time. According to Prospera a grade of more than four should be interpreted as "excellent".
2 The survey was conducted between 16 November and 29 December 2015. Seven primary dealers and 54 investors were polled. The response rate was 89 per cent.
2.1 Liquidity and infrastructure
Usually the demand for government securities is based on their low credit risk and good liquidity.
The low credit risk follows from the fact that they are securities issued by the State. In addition, Sweden has strong central government finances, its own currency and its own central bank. Their attractiveness is also due to a great extent to the liquidity that the government securities market can offer. What is meant here is the possibility of selling or buying large volumes of government securities without any appreciable effect on their price.
To attract as broad a range of investors as possible the Debt Office is working actively to promote the liquidity and infrastructure of the government securities market. This is mainly done by:
active market and debt management measures with switches and repo facilities
concentrating borrowing to a limited number of benchmark bonds
maintaining several effective sales channels, partly with the aid of a system of primary dealers.
Market and debt commitment
To reduce the risk of the Debt Office's primary dealers in government securities not obtaining bonds or bills that are particularly sought after, the Debt Office offers repos in government securities.
The knowledge that the repos are available makes it easier for primary dealers to quote prices, which then benefits liquidity of the market.
From experience, repos of government securities are viewed as important when primary dealers rank Debt Office activities. It is therefore satisfying that the Debt Office has been given the excellent grade of 4.3.
The Debt Office also offers switches of less liquid government securities for more liquid issues. New 3,6
3,8 4,0 4,2 4,4
2004 2006 2008 2010 2012 2014
Primary dealers Swedish investors International investors Grade
2 Debt management strategies
This chapter describes the management strategies intended to achieve the Debt Office’s intermediate objectives. The chapter also reports what primary dealers and investors think about how the Debt Office is implementing these strategies.
bonds are normally introduced using switches so that the new bonds quickly gain good liquidity.
In 2015 the Debt Office offered larger switch volumes of inflation-linked bonds than in the past, partly in connection with the introduction of two new bonds, but mainly after liquidity deteriorated following rumours that the Riksbank was going to buy inflation-linked bonds for monetary policy purposes.
The action taken by the Debt Office in the inflation- linked bond market can be read off in the results of this year’s survey. The grade for switches of inflation-linked bonds through the standing facility was 4.0 for 2015 compared with 3.5 for the preceding year. This is a significant improvement.
Section 6.2 discusses work on market commitment in the inflation-linked bond market in more depth.
Few benchmark bonds
Borrowing in government bonds and inflation-linked bonds is usually concentrated to a few benchmark bonds in order to make sure that the outstanding volume of these bonds is sufficient to ensure good liquidity. This is accentuated when the central government debt shrinks. When issue volumes are large, there is less need to concentrate borrowing.
Government bonds are the Debt Office’s most important instrument and when the borrowing requirement is small, issues of these bonds are given priority ahead of other borrowing so as to maintain liquidity in the government bond market.
The Debt Office endeavours to maintain a relatively even maturity profile in its bond stock. By offering bonds with both short and long maturities the Debt Office is able to attract different types of investors.
If the maturity profile is even, the refinancing risk is also smaller since this means that only a small part of the central government debt matures each year.
Effective sales channels
The Debt Offices tries to maintain many, effective sales channels. A system of primary dealers is a guarantee for well-functioning infrastructure in the Swedish government bond market. It also
contributes to better liquidity and good possibilities of borrowing large volumes in a possible crisis situation in the future.
When the Debt Office sells bond in foreign currencies, it does so through ‘syndication’. This
means that the Debt Office engages a group of banks, a syndicate, to handle the sale.
Liquidity has deteriorated but is still good
The Debt Office can only create the conditions for good liquidity in the market, but has small possibili- ties of influencing liquidity directly. This year's survey shows that both investors and primary dealers consider that liquidity in the Swedish government bond market has deteriorated. The reasons usually stated are the low interest rate and new regulations. In addition, the Riksbank's purchases of these bonds have reduced the outstanding stock of securities available for trading (see the box The Riksbank's purchases of
government bonds).
However, the liquidity of government bonds is still regarded as good even though traded volumes are lower. Swedish investors give good grades for both liquidity and price transparency, while primary dealers and international investors are slightly more negative.
The survey shows that liquidity is also thought to have deteriorated in the market for inflation-linked bonds. Here, however, liquidity is always lower than for nominal bonds. Essentially this is because investors do not have the same interest in active trading in inflation-linked bonds. Nor is there a developed market in derivative instruments here. As already mentioned, the Debt Office made special efforts during the year to support liquidity by increasing its commitment to switch between different inflation-linked bonds.
Liquidity in the market for T-bills has deteriorated further from an already low level. The outstanding stock of bills has been more than halved in five years, and investors are increasingly using other instruments to invest money in the short term.
2.2 Transparency, predictability and clear communication
To create an attractive market for government securities, these securities have to be managed with transparency and predictability. This means that all communication with the market should be as open and consistent as possible. If counterparties and investors have a good knowledge of the Debt Office's issue plans and are aware of how the Debt Office responds to external change, this creates stable rules and less uncertainty.
The Debt Office tries to achieve the objectives of transparency, predictability and clarity by:
maintaining good investor relations
regularly publishing forecasts of the borrowing requirement for coming years handling borrowing in a consistent way with clear principles
having clear communication both in written documents and in contacts with investors and counterparties
providing detailed information about borrowing and the central government debt on its website: riksgalden.se.
Investor relations
Ultimately a broad investor base results in lower interest costs. Working for good investor relations is therefore one of the Debt Office’s most important management strategies.
Primary dealers have the most active role in sales of government securities. They provide investors with analysis and information about the Debt Office's issues. However, the Debt Office does not rely solely on information given via its primary dealers. It is important that investors can receive, without intermediaries, the information they consider they need. The Debt Office therefore meets Swedish and foreign investors at both personal meetings and large conferences. The Debt Office also holds investor meetings when new borrowing forecasts are presented.
In 2015 there were visits to investors in Europe, South America and the US. Representatives of the Debt Office participated as speakers at several seminars and conferences in Sweden and abroad.
In this year’s survey the Debt Office got a grade of 3.5 for contacts with investors. So there is potential for improvement here. Swedish investors gave a
The Riksbank’s purchases of government bonds
In 2015 the Riksbank purchased nominal government bonds in order to press down the general level of interest rates, which is, in turn, intended to contribute to higher inflation.
The Riksbank bought bonds with a nominal value of SEK 135 billion, corresponding to 22.4 per cent of the outstanding stock of nominal
government bonds. In 2016 the Riksbank intends to purchase bonds for a further SEK 65 billion and is expected, in June 2016, to hold a third of the outstanding stock.
The Debt Office finances the central government deficit and manages the government debt. The Riksbank’s purchases of government bonds do not alter either the financing requirements of central government or the direction of the management of the debt.
However, the Riksbank’s purchases do change the size of the outstanding stock of bonds traded actively on the market. Therefore it cannot be rule out that liquidity deteriorates as a result of the Riksbank’s purchases of government.
The Riksbank’s purchases of nominal government bonds and outstanding stock, 31 December 2015
2.8%
27.9% 27.7% 30.3%
24.8% 20.8%
30.9%
23.3%
0.0% 3.4%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 10 20 30 40 50 60 70 80 90 100
1050 1051 1052 1047 1054 1057 1058 1059 1056 1053
Outstanding stock Purchases by the Riksbank Fraction (right scale)
SEK billion Fraction
grade of 3.3 for the reason that the Debt Office must listen more to the market as a result of the poorer liquidity. During the year the Debt Office has increased its attentiveness to the market, in part by offering two bonds at most auctions to a greater extent than before (see section 6.1).
For the second year in a row the interest of investors in having closer contacts with the Debt Office increased. Most prefer personal meetings, but many also show interest in the Debt Office’s investor meetings. Almost 80 per cent of Swedish investors want to have contact once a quarter or once every six months. In general, the international investors are satisfied with one contact per year.
Predictability and clear communication The report Central government borrowing – forecast and analysis is important for the Debt Office to communicate its plans to all market players at the same time. This report is published three times a year and describes in detail how the Debt Office intends to finance the central
government debt in the coming two years.
The issue plan is based on a forecast of the net borrowing requirement and information about refinancing needs and interest payments on the existing debt. The Debt Office adapts borrowing so as to fulfil the conditions in the government guide- lines. The issue plan and the forecast for the net borrowing requirement normally remain in place until the next report is published. Good forecasts are necessary in order to be able to act in a predictable way (see chapter 4 for a description of the forecasts).
Sweden's Central Government Debt is published on a monthly basis and contains detailed informa- tion about the composition of the debt.
For 2015 the Debt Office was given a grade of 4.1 for communication about borrowing requirements and financing and a grade of 4.0 for clear and consistent action. These grades are marginally poorer than those for 2014.
The website riksgalden.se
During the year the Debt Office has expanded the information about borrowing and the debt on its website. It is now possible to gather information interactively on the website for downloading.
The Debt Office’s website remains the most important channel for information about the
borrowing requirement and financing, auction terms and auction outcomes. 92 per cent of the primary dealers use the Debt Office website. Among Swedish and international investors the corre- sponding figures are 97 and 66 per cent. Almost 100 per cent of those who visit the website con- sider they find the information they are looking for.
2.3 Confidence in the Debt Office
Prospera’s summary of this year’s results reads as follows: ”The rating among primary dealers can be characterised as very high. Among Swedish and international investors the rating is lower, but can still be described as good.”
The survey shows that confidence in the Debt Office remains high and that, as before, the Debt office is given a high ranking for its transparency compared with other Debt Offices.
When respondents were asked to rank the importance of the Debt Office's strategies, market commitment through repos came top. Table 1 shows the five most important demands made by market players.
Table 1 The most important requirements
Requirements Grade
Market commitment through repos 4.8 Communication about borrowing requirements and
financing 4.7
Clear and consistent actions 4.5
Information about volumes and other conditions
concerning government securities 4.5 Market commitment through switches of inflation-linked
bonds 4.3
When the same market players graded the ability of the Debt Office to implement the strategies, the ranking was the same, see table 2. A reasonable conclusion from this is that the Debt Office is focusing on the rights things.
Table 2 The Debt Office’s main strengths
Strengths Grade
Market commitment through repos 4.3 Communication about borrowing requirements and
financing 4.1
Clear and consistent actions 4.0
Information about volumes and other conditions
concerning government securities 4.0 Market commitment through switches of inflation-linked
bonds 4.0
The overall objective of the management of the central government debt is to minimise the long- term cost of the debt while taking account of risk.
However, evaluating whether the overall objective has been met is a complicated matter. Since the Debt Office is the dominant borrower in Swedish kronor, there are no natural comparisons against which to evaluate the strategy. Comparisons with hypothetical alternatives are also difficult to make, partly because the size of the central government debt means that interest rates are affected by the borrowing strategy actually implemented.
The evaluation of the cost of the government debt therefore has to make use of several measures. The risks in the management of the debt are described in both quantitative and qualitative terms.
3.1 Costs
The cost of the central government debt can be derived from the cash flows (payments) that the debt generates in the future. In the case of the nominal krona debt, the cash flows are solely due to the interest rate at which the loan was raised. In the case of the inflation-linked debt, the cash flows also depend on inflation, and for the foreign currency debt cash flows are affected by exchange rates.
It is difficult to calculate the cost of the central government debt in a simple way using a single figure. The Debt Office therefore reports the cost using three different measures:
interest payments
period cost
average issue yield
Interest payments show how much is paid in a single year. The measure is linked to the cash flows.
This means that when an instrument is issued at a premium (in relation to its nominal value) the difference between the actual and the nominal
amount is counted as revenue immediately instead of being spread over the term of the loan. In the same way, the cost is charged immediately when an instrument is issued at a discount. Nor are the outcomes of repurchases accrued.
The measure period cost includes virtually all cash flows.3 Assumptions are made about future inflation and exchange rates in order to forecast cash flows.
These are then related to the actual (rather than nominal) amounts that have been borrowed and are spread over the term of the loan.
Average issue yield is a weighted average of the yields at which currently outstanding loans were raised. This measure does not take account of either inflation compensation or changes in exchange rates and only gives a snapshot of the average borrowing cost.
Interest payments
Interest payments amounted to SEK 22 billion in 2015. This is almost SEK 19 billion more than for 2014, which is chiefly explained by the temporary reduction in interest payments in 2014 due to high premiums and foreign currency gains. In 2015 the effects operated in the other direction instead, increasing interest payments. For example, currency exchange losses of around SEK 5 billion were incurred in CHF as a result of the Swiss central bank’s decision to abandon the link to the euro in January 2015. Currency exchange losses were also incurred in USD and GBP in 2015.
Interest payments were also affected by the
payment of around SEK 5 billion in accrued inflation compensation when inflation-linked bond SGB IL 3105 matured in December 2015.
Otherwise the lower levels of market interest rates have helped to reduce central government interest payments. This is seen both through the positive cash flows generated by the interest rate swaps
3 Cash flows attributable to FX Forwards are not included in the measure.
3 Cost and risk
This chapter gives an account of various measures of cost and risk in the central government debt. It also describes how the composition and maturity of the debt have been steered on the basis of the guidelines that applied in 2015.
entered into by the Debt Office and through the decrease in the coupon rate (interest payments) as old loans are replaced by new ones.
Period cost
The period cost contains all the components that affect the cost of the central government debt:
interest, inflation compensation and exchange rate fluctuations.
The cost for a particular period is calculated by relating the amounts borrowed to the cash flows that will be paid in the future, taking account of when in time the payments will be made. For the inflation-linked debt, the foreign currency debt and swaps future cash flows are not known. It is there- fore assumed that future inflation will follow the break-even inflation, i.e. the inflation derived from the market pricing of inflation-linked bonds in relation to nominal bonds. Future exchange rates and short-term interest rates are assumed to stay at the same level as at the end of 2015.
Over time these assumptions are replaced by the actual outcomes, which mean that earlier costs can be revised. The earlier cost is also affected when the Debt Office buys back government securities since gains or losses are then realised and replace the assumption that all loans are held to maturity.
The cost of every instrument is spread evenly over its term. For a foreign currency loan, for example, this means that day-to-day fluctuations in exchange rates do not affect the cost if they are not realised.
It is only the exchange rates that affect the actual cash flows that have an effect on the cost.
Figure 2 Period cost for all types of debt4
Figure 2 shows the period cost for all types of debt.
For the nominal krona debt the cost was 1.5 per cent in 2015. The cost has shown a falling trend for
4 Here the calculation of the cost of the foreign currency debt includes foreign currency bonds and the legs of swaps that are denominated in foreign currency.
a long period, reflecting the general decline in interest rates.
In 2015 the cost of the inflation-linked debt was 2.6 per cent. The cost of the inflation-linked debt has also shown a falling trend for a long period as real interest rates have fallen. Moreover, inflation has been unusually low, as illustrated in figure 3.
Figure 3 Inflation rate, annual CPI change
The cost of the inflation-linked krona debt has been higher than the cost of the nominal krona debt for a long time. This is mainly because the maturity of the inflation-linked krona debt is longer, so the inflation- linked debt stock contains a larger share of loans raised long ago at much higher interest rates.
For the foreign currency debt the cost was 0.3 per cent in 2015. One explanation of the low cost is the very short maturity of the foreign currency debt. As a result, low market rates impact quickly on cost.
But the cost calculation for the foreign currency debt is not complete since FX forwards are not included in the cost measure.
Figure 4 Annual change in the krona exchange rate in relation to the foreign currency debt
The krona exchange rate is of great importance for the cost of the foreign currency debt. In 2015 the krona weakened against most currencies but strengthened against the euro, which accounts for the bulk of the foreign currency exposure. Overall the krona weakened by about one per cent in relation to the foreign currency debt, see figure 4.
-2 0 2 4 6 8
2002 2004 2006 2008 2010 2012 2014 Nominal krona debt Inflation-linked debt Foreign currency debt
Per cent
-1 0 1 2 3 4
2002 2004 2006 2008 2010 2012 2014 Per cent
-10 -5 0 5 10 15 20
2002 2004 2006 2008 2010 2012 2014 Per cent
How quickly exchange rate fluctuations impact on the cost of the foreign exchange debt depends on how the debt was created. A large part has arisen by swapping krona borrowing for foreign currency (see the box on page 28 for details). The maturity of these swaps is relatively long, and this means that exchange rate fluctuations will affect the cost gradually over a long period of time.
In the case of the part of the debt that comes from borrowing in foreign currencies, exchange rate fluctuations have a quicker impact.
Figure 5 Period cost for various government securities
Figure 5 shows the period cost for various govern- ment securities. This cost has been lower and more varied for T-bills since their maturity is shorter so that interest changes have a quicker impact. In 2015 the cost was negative, -0.3 per cent. For government bonds the cost was 2.7 per cent, and it was 2.6 per cent for inflation-linked bonds.
Average issue yield
For the nominal krona debt the average issue yield was 0.5 per cent at the end of 2015, a decrease from 1.1 per cent in 2014, see figure 6. The low yield is due to the general decline in interest rates.
Figure 6 Average issue yield
For the same reason, the average real issue yield also fell. At the end of 2015 it was 1.5 per cent,
which is 0.3 percentage points lower than a year previously.
The average issue yield for the foreign currency debt adjusts quickly to the fall in market rates since the maturity of the debt is short. At the end of 2015 the issue yield was -0.1 per cent. This is 0.1 per- centage points lower than a year previously.
It should be noted that the issue yields for different types of debt are not fully comparable. In the cases of the inflation-linked and foreign currency debt, the whole cost is not shown since inflation compen- sation and exchange rate fluctuations are not included in the calculations. In addition, the debt types have different maturities.
3.2 Risks
Amended guidelines
Ahead of 2015 the Government decided to change the principles for the calculation of maturities and shares in the steering of the central government debt. The change means that maturity is to be calculated as duration and that the shares are calculated using nominal amounts at the current exchange rate including inflation compensation.
As a result of the new principles, changes were made to the benchmarks for the maturity of the nominal and inflation-linked krona debt and for the share of inflation-linked bonds in the debt. But the actual composition of the debt was not altered.
Another change to the guidelines was that the benchmark for the currency exposure of the central government debt was replaced with a target to reduce the currency exposure by no more than SEK 30 billion per year (the box on page 3 describes the background to that decision).5
Maturity
According to the guidelines that applied at the start of the year, the duration of the nominal krona debt consisting of instruments with a maturity of up to twelve years was to be between 2.3 and 2.8 years.
This was a translation of the old measure of interest rate refixing period (IRRP) adjusted for the increase in maturity that followed from the decrease in foreign currency exposure.
5 Changes in the SEK exchange rate are excluded when calculating the exposure.
-2 0 2 4 6 8
2002 2004 2006 2008 2010 2012 2014 Government bonds T-bills
Inflation-linked bonds Per cent
-2 0 2 4 6
2003 2005 2007 2009 2011 2013 2015
Nominal krona debt Inflation-linked debt Foreign currency debt
Per cent
After the government decision on new guidelines in November 2014 there was a sharp fall in market rates. Since duration is a measure that is affected by market rates, the relationship between duration and the old measure of IRRP changed. In March 2015 the Government therefore decided to raise the duration by 0.3 years to between 2.6 and 3.1 years. This decision meant that the Debt Office was able to follow the issue plan that it had in place.
Figure 7 shows the duration of the central
government debt as a whole. To simplify historical comparisons the old measure of IRRP is also shown. The figure shows how the measures have converged as interest rates have fallen.
Figure 7 Duration and IRRP for the whole central government debt
For nominal instruments with a maturity in excess of twelve years the Government decided that the long-term benchmark for the outstanding volume was to stay at SEK 70 billion.
The duration of the inflation-linked krona debt was to be between six and nine years, while it was to be 0.125 years for the foreign currency debt.
The duration of the nominal krona debt varies because central government payments are uneven over time. On days with large deficits the Debt Office borrows more at short maturities and vice versa. Borrowing is planned so as to hold the duration between 2.6 and 3.1 years. However, it is not the intention of the interval to force the Debt Office to make transactions so as to steer maturity if exogenous events have resulted in the duration being longer or shorter than intended. The Government has therefore decided that temporary deviations from the maturity intervals are permitted.
Figure 8 shows the duration and IRRP for the nominal and inflation-linked krona debt. In 2015 the average duration of the nominal krona debt was 2.8 years, while it was 6.8 years for the inflation-linked
krona debt. The average duration of the foreign currency debt was only 0.1 years.
Figure 8 Duration and IRRP for nominal and inflation-linked debt
The conditions for steering duration are different for the different types of debt. In the international capital market the Debt Office is a small participant, and it is easy to steer the duration of the foreign currency debt using interest rate derivatives. The duration of the nominal krona debt can also be steered to some extent by varying the volume of interest rate swaps. But when it comes to the inflation-linked debt, there is no possibility of using derivatives to adjust duration. The duration of the inflation-linked debt is therefore determined solely by the maturity of the outstanding bonds.
Inflation-linked krona debt
According to the guidelines the inflation-linked krona debt was to be 20 per cent of the total debt.6 On average it was 18.6 per cent in 2015. The benchmark of 20 per cent is a long-term figure and the share is hard to steer in the short term. For example, it falls a great deal when a bond loan matures. This explains why the inflation-linked share fell sharply at the end of 2015, see figure 9. The figure also shows the old composition measure in order to facilitate comparisons over time.
Figure 9 Inflation-linked share
6 This share is calculated on the basis of the nominal amount of the inflation-linked krona debt including accrued inflation compensation.
2 3 4 5
2010 2011 2012 2013 2014 2015 2016
Duration Interest rate refixing period Years
4 6 8 10
2 3 4 5
2010 2011 2012 2013 2014 2015 2016
Nominal DUR Nominal IRRP
Inflation-linked DUR Inflation-linked IRRP Nominal debt, years Inflation-linked debt, years
15 20 25 30
2010 2011 2012 2013 2014 2015 2016
Current measure Old measure Per cent
Foreign currency debt
The foreign currency debt consists of several currencies. In 2015 the Debt Office reduced the foreign currency debt by exchanging JPY 318 billion, corresponding to SEK 22 billion calculated using the year-end exchange rate. The reason why the reduction was made exclusively in JPY was that this was judged to be the most effective way of reducing the risk in the foreign currency debt. The exchanges were made at an even pace over the year to avoid making them at a few unfavourable points in time.
Figure 10 Foreign currency debt at current exchange rate
The size of the reduction of the total foreign currency debt during the year can be measured in several ways. Measured at current exchange rates it decreased by about SEK 25 billion in 2015, see figure 10. If exchange rate fluctuations are excluded according to the method used by the Debt Office, the reduction is instead SEK 28 billion.
Refinancing risk
Refinancing risk is the risk that loans reaching maturity can only be replaced with new loans at very high costs or, in the extreme case, cannot be refinanced at all. The refinancing risk is generally held to be higher the larger the loans that are maturing in the immediate future. To some extent this is a simplification since a bond that is maturing seldom needs to be financed right away by issuing another bond. With long-term issue planning and small issue volumes at regular auctions refinancing is spread over a long period of time and old bond loans are often replaced before maturing.
Bond maturities work just like other government payments. The net amount of daily payments varies and can total SEK 100 billion on certain days. The refinancing risk is therefore only part of what is usually called liquidity risk (or financing risk), which is to do with the possibility of managing payments more generally.
Nevertheless the refinancing risk should be controlled. The ways in which the Debt Office tries to limit it is by working to maintain an even maturity profile for government and inflation-linked bonds and by contributing to the establishment of a well- functioning market in government securities.
Figure 11 Maturity profile in December 2015
Figure 11 shows the maturity profile of the central government debt, i.e. the size of the sums in the outstanding debt that mature in every single year.
The figure also shows claims that mature in the form of on-lending to the Riksbank. These claims match the bond loans in foreign currency to a great extent.
Normally a large part of the debt falls due in the coming year. This is because of the comprehensive liquidity management and the handling of seasonal fluctuations in the net borrowing requirement. At the end of the year the short-term maturities are at their largest level since a large part of central government payments take place in the month of December. These payments are financed initially by money market borrowing and are then gradually replaced by borrowing at longer maturities. Chapter 7 gives a more detailed description of money market borrowing.
Otherwise the maturity profile is relatively even up until ten years, and this influences future refinancing risks as time passes. An uneven maturity profile could give rise to greater risks in the future when large maturities must be refinanced. However, what is most important for risk at a given point in time is what is maturing in the near future.
Figure 12 shows two measures of refinancing risk.
The first shows what share of the debt will mature within twelve months. The second is defined by giving a loan maturing immediately the value of one while a loan that never matures is given the value of zero. All loans are then weighted together and this results in a figure between zero and one for the total central government debt.
150 160 170 180 190 200
2010 2011 2012 2013 2014 2015 2016 SEK billion
-100 0 100 200 300 400
2016 2021 2026 2031 2036
Liquidity mgmt, CP, collateral T-bills
Retail borrowing Bonds in foreign currency Inflation-linked bonds Government bonds On-lending SEK billion
Figure 12 Refinancing risk
The measures show the same pattern even though the second one does not have the sharp break at 12 months. The figure also shows a clear seasonal variation on account of the extensive money market borrowing around the turn of each year.
The refinancing risk decreased during the financial crisis in 2008/09. Unlike many other government borrowers, the Debt Office was able to reduce the share of short-term financing when the borrowing requirement increased. Since then the refinancing risk has gone up again on account of an increase in short-term borrowing. The share of the central government debt financed in the money market is illustrated in figure 13. In 2015 that share was 16 per cent on average.
Figure 13 Share of debt financed in the money market
Counterparty risk
Counterparty risk means the risk that the counterparty in a transaction cannot fulfil its obligations to pay or deliver collateral.
Counterparty risks arise both when placing surpluses in liquidity management and when the Debt Office enters into derivative transactions without central counterparty clearing. The risks in investments and derivative transactions are handled
differently, but in both cases the Debt Office sets minimum requirements concerning the credit rating of its counterparty.
Through liquidity management the Debt Office borrows or places funds on a daily basis so as to guarantee that central government can make its payments at as low a cost as possible. To handle the counterparty risk for these placements there are limits based on the counterparty’s credit rating that restrict the maximum exposure and maturity.
Derivative transactions are used to steer the maturity of the central government debt and to take positions in day-to-day position-taking.
Transactions can either be closed bilaterally or through central counterparty clearing.
To enable the Debt Office to handle derivative instruments that are not cleared centrally, the Debt Office and its counterparty draw up an ISDA agreement with a downgrade clause and a Credit Support Annex (CSA). The CSA agreement contains thresholds that govern the maximum permitted exposure to the counterparty. If the value of the exposure exceeds these thresholds, the counterparty must provide the Debt Office with collateral. This collateral provides protection in the event that the counterparty is unable to meet its commitments. The size of the threshold depends on the counterparty’s credit rating. The Debt Office's ISDA/CSAs are bilateral in the sense that the Debt Office not only accepts but also provides collateral if the counterparty has a positive exposure in relation to the Debt Office.
Regulations of this type are defined and updated in the context of the Debt Office’s Financial and risk policy.
In 2015 there were no material events that had an effect on the Debt Office’s credit risk. But
counterparty risk was affected by the fact that a number of counterparties were given higher or lower credit ratings. Two counterparties had their credit rating reduced to below the Debt Office’s minimum permitted level (A-). There were outstanding deals with them at the time of the reduction. After weighing the cost of closing the outstanding deals prematurely against the risk in retaining the deals, the Debt Office chose to retain them.
0,35 0,40 0,45 0,50 0,55
10 20 30 40 50
2002 2004 2006 2008 2010 2012 2014 Measure 1: Debt maturing within twelve months (left) Measure 2: Refinancing risk (right)
Per cent
0 10 20 30
2003 2005 2007 2009 2011 2013 2015 Per cent
The net central government borrowing requirement is the single most important factor for the
development of the central government debt. The Debt Office therefore makes detailed forecasts of the net borrowing requirement, in both the short and the long term. The purpose of these forecasts is to create conditions for stable issue plans and effective liquidity management. In somewhat simplified terms, the annual forecasts steer long- term bond borrowing and the monthly forecasts steer short-term borrowing in commercial paper and T-bills, while the daily forecasts affect day-to- day liquidity management.
Annual forecasts for 2015
In autumn 2013 the forecast of the net borrowing requirement in 2015 was SEK 18 billion. Then it was assumed that 2015 would be characterised by a strong recovery of the global economy and that the Swedish economy would therefore gain momentum from international developments. But this did not happen and up until autumn 2014 the Debt Office underestimated the net borrowing requirement.
Table 3 Annual forecasts for 2015 SEK billion
2014 Dec
2015 Feb
2015 Jun
2015 Oct
Out- come Primary borrowing
requirement 30 52 46 22 11
of which on-lending 5 10 10 10 10 of which income from sales 0 0 0 0 0 Interest on the central
government debt 21 29 25 24 21
Net borrowing
requirement 51 80 71 45 33
Net borrowing req., excl.
on-lending and sales
income 46 71 61 36 23
In late 2014 and early 2015 there was a downward revision of the assumptions about global growth.
Geopolitical tensions in the Middle East and Ukraine and weaker developments in the emerging economies dampened expectations for the future.
This led to the upward revision of forecasts of the net borrowing requirement. In the first central government borrowing report (2015:1) the forecast was SEK 80 billion, see table 3.
Global growth in 2015 was moderate. Despite this the Swedish economy showed surprisingly strong growth, driven by domestic demand. The forecasts of the net borrowing requirement were adjusted downwards gradually in the second half of the year, mainly because tax payments exceeded
expectations. But the net borrowing requirement was still over-estimated. The deviation between the final forecast and the outcome was just under SEK 13 billion.
Comparison with forecasts by other government agencies
In addition to the Debt Office, the National Institute of Economic Research (NIER), the National Financial Management Authority (ESV) and the Ministry of Finance also make forecasts of the central government net borrowing requirement.
These agencies and the Ministry of Finance have different principles for dealing with sales income in forecasting contexts. To simplify comparisons this income has been excluded from figure 14, see below. All the forecasters had a similar profile in their forecasts. In autumn 2013 and in early 2014 the net borrowing requirement was underestimated.
After that the forecasts were revised strongly upwards on account of weak global prospects for growth.
4 Borrowing requirement forecasts
This chapter evaluates the Debt Office’s forecasts of the net borrowing requirement on an annual, monthly and daily basis. A comparison is also made with forecasts of the annual borrowing requirement made by other government agencies.
Figure 14 Comparison with other agencies
In the second half of 2015 the forecasts were gradually revised downwards. Nevertheless all these forecasters overestimated the net borrowing requirement in their final forecasts for the year.
Monthly forecasts
At present the Debt Office is the only government agency to publish monthly forecasts of the net borrowing requirement. This means that comparisons cannot be made with other forecasters. The Debt Office follows up the precision of its monthly forecasts using the measure of Root Mean Square Error, RMSE.7 Figure 15 Deviation in monthly forecasts
according to RMSE, 2011–2015
Figure 15 shows the development of RMSE since 2011. It can be seen that RMSE decreased in 2015 compared with 2014.
7 RMSE is defined as
√(𝑒12+ 𝑒22+ 𝑒32+ 𝑒42)/4
where 𝑒𝑡 is the forecasting error (outcome of the net borrowing requirement as a proportion of the last published forecast) for month t.
The Debt Office updates its forecast every four months, so the RMSE is based on the forecasting error in the past four months.
On occasion the net borrowing requirement has been affected strongly by the Debt Office’s on- lending and income from sales of state-owned enterprises. For this reason an adjusted RMSE excluding these items is also presented. Expressed in terms of this measure, the precision of forecasts has been relatively even during the period.
Daily forecasts
The Debt Office also makes daily forecasts for the coming six months so as to plan its liquidity management. Unlike the annual and monthly forecasts, the daily forecasts are updated
continually as new information becomes available.
The average deviation per day was SEK 131 million in 2015. Figure 16 shows daily outcomes and forecasts for central government’s net primary borrowing requirement when the Debt Office’s net lending to all other agencies has been excluded.
The distance to the 45 degree line is the deviation in the forecast for each day. If the forecasts did not have any information value, the points in the figure would be distributed randomly.
Figure 16 Forecast and outcome on a daily basis
The largest deviations at the daily level concerned tax income. Compared with previous years the pattern of payments for tax income changed in 2015. A larger share of taxes was paid in early in the month and a considerably smaller share than before was paid on the due date.
This change in the payment pattern could be an effect of the low level of interest rates. Since tax accounts gave tax-free interest of 0.56 per cent in 2015, there was an incentive to bring tax payments forward. Nor can the possibility that companies have used their tax account to place liquid funds be ruled out.
0 25 50 75 100
aug-13 okt-13 dec-13 feb-14 apr-14 jun-14 aug-14 okt-14 dec-14 feb-15 apr-15 jun-15 aug-15 okt-15 dec-15
ESV Ministry of Finance
Outcome NIER
Riksgälden SEK billion
05 1015 2025 3035 4045
RMSE Adjusted RMSE
SEK billion
-60 -40 -20 0 20 40 60
-60 -40 -20 0 20 40 60
Outcome, SEK billion
Forecast, SEK billion
At year-end the central government debt was SEK 1 403 billion. The net borrowing requirement in 2015, i.e. the deficit in the central government budget, was SEK 33 billion, which is SEK 40 billion lower than for 2014. In contrast, gross borrowing increased by SEK 46 billion and reached SEK 479 billion. The increase is due to large maturities of money market loans in 2015, as shown in table 4.
Table 4 Gross borrowing requirement
SEK billion 2014 2015
Net borrowing requirement 72 33
Business day adjustment etc.1 –4 0
Retail market & collateral, net2 3 31
Maturities, money market3 180 256
T-bills 94 88
Commercial paper 39 124
Liquidity management instruments 47 44 Maturities, switches and buybacks, capital
market 182 160
Government bonds 89 75
Inflation-linked bonds 8 31
Foreign currency bonds 86 54
Total borrowing requirement, gross4 433 479
1 Adjustment for difference between settlement date and business date.
2 Net change in retail market borrowing and collateral.
3 Initial stock maturing within 12 months.
4 Refers to borrowing requirement in the institutional market.
Interest rates fell to record low levels in 2015. Ten- year government bonds were issued at an average yield of 0.48 per cent, compared with 1.35 per cent in 2014.
The Debt Office continued to give priority to borrowing in government bonds, which is the most important borrowing instrument in the long term. A total of SEK 86 billion was issued compared with SEK 77 billion in 2014 (see table 5).
In the case of inflation-linked bonds the planned issue volume was SEK 18 billion. The amount issued reached SEK 17 billion since some issues were not fully subscribed.
Table 5 Total gross borrowing
SEK billion 2014 2015
Money market, borrowing 256 284
T-bills 88 141
Commercial paper 124 87
on behalf of central government 117 78 on-lending to the Riksbank 6 9 Liquidity management instruments 44 56
Capital market, borrowing 177 194
Government bonds 77 86
Inflation-linked government bonds 17 17
Foreign currency bonds 84 91
on behalf of central government 25 38 on-lending to the Riksbank 59 53
Total gross borrowing 433 479
During the year the Debt Office raised bonds in foreign currencies corresponding to SEK 91 billion, compared with SEK 84 billion in the preceding year. The increase is primarily due to an increase in borrowing on behalf of central government to SEK 38 billion from SEK 25 billion. The remainder, SEK 53 billion, related to refinancing of loans to the Riksbank. A further SEK 9 billion of the Riksbank’s maturing loans was refinanced with commercial paper.
The Debt Office was able to borrow in foreign currencies on good terms during the year despite periods of uncertainty about US monetary policy and volatility in the interest and stock market.
The outstanding stock of T-bills amounted to SEK 141 billion at the end of 2015. This is SEK 53 billion more than the preceding year. The Debt Office uses T-bills both in its regular debt management and in its liquidity management. The increase in the T-bill stock in 2015 is because the Debt Office issued more as part of its liquidity management.
5 Summary of borrowing
This chapter gives a summary overview of the Debt Office’s borrowing requirement and borrowing in 2015.
Even though the budget deficit decreased, gross borrowing increased by SEK 46 billion.